Archival Storage is a solution to decouple growing storage capacity from compute capacity. Nodes with higher density and less expensive storage with low compute power are becoming available and can be used as cold storage in the clusters. Based on policy the data from hot can be moved to the cold. Adding more nodes to the cold storage can grow the storage independent of the compute capacity in the cluster.
The frameworks provided by Heterogeneous Storage and Archival Storage generalizes the HDFS architecture to include other kinds of storage media including SSD and memory. Users may choose to store their data in SSD or memory for a better performance.
The first phase of Heterogeneous Storage (HDFS-2832) changed datanode storage model from a single storage, which may correspond to multiple physical storage medias, to a collection of storages with each storage corresponding to a physical storage media. It also added the notion of storage types, DISK and SSD, where DISK is the default storage type.
A new storage type ARCHIVE, which has high storage density (petabyte of storage) but little compute power, is added for supporting archival storage.
Another new storage type RAM_DISK is added for supporting writing single replica files in memory.
A new concept of storage policies is introduced in order to allow files to be stored in different storage types according to the storage policy.
We have the following storage policies:
More formally, a storage policy consists of the following fields:
When there is enough space, block replicas are stored according to the storage type list specified in #3. When some of the storage types in list #3 are running out of space, the fallback storage type lists specified in #4 and #5 are used to replace the out-of-space storage types for file creation and replication, respectively.
The following is a typical storage policy table.
Policy ID | Policy Name | Block Placement (n replicas) | Fallback storages for creation | Fallback storages for replication |
---|---|---|---|---|
15 | Lazy_Persist | RAM_DISK: 1, DISK: n-1 | DISK | DISK |
12 | All_SSD | SSD: n | DISK | DISK |
10 | One_SSD | SSD: 1, DISK: n-1 | SSD, DISK | SSD, DISK |
7 | Hot (default) | DISK: n | <none> | ARCHIVE |
5 | Warm | DISK: 1, ARCHIVE: n-1 | ARCHIVE, DISK | ARCHIVE, DISK |
2 | Cold | ARCHIVE: n | <none> | <none> |
Note that the Lazy_Persist policy is useful only for single replica blocks. For blocks with more than one replicas, all the replicas will be written to DISK since writing only one of the replicas to RAM_DISK does not improve the overall performance.
When a file or directory is created, its storage policy is unspecified. The storage policy can be specified using the “storagepolicies -setStoragePolicy” command. The effective storage policy of a file or directory is resolved by the following rules.
If the file or directory is specificed with a storage policy, return it.
For an unspecified file or directory, if it is the root directory, return the default storage policy. Otherwise, return its parent’s effective storage policy.
The effective storage policy can be retrieved by the “storagepolicies -getStoragePolicy” command.
dfs.datanode.data.dir - on each data node, the comma-separated storage locations should be tagged with their storage types. This allows storage policies to place the blocks on different storage types according to policy. For example:
The default storage type of a datanode storage location will be DISK if it does not have a storage type tagged explicitly.
A new data migration tool is added for archiving data. The tool is similar to Balancer. It periodically scans the files in HDFS to check if the block placement satisfies the storage policy. For the blocks violating the storage policy, it moves the replicas to a different storage type in order to fulfill the storage policy requirement. Note that it always tries to move block replicas within the same node whenever possible. If that is not possible (e.g. when a node doesn’t have the target storage type) then it will copy the block replicas to another node over the network.
Command:
hdfs mover [-p <files/dirs> | -f <local file name>]
Arguments:
-p <files/dirs> | Specify a space separated list of HDFS files/dirs to migrate. |
-f <local file> | Specify a local file containing a list of HDFS files/dirs to migrate. |
Note that, when both -p and -f options are omitted, the default path is the root directory.
List out all the storage policies.
Command:
hdfs storagepolicies -listPolicies
Arguments: none.
Set a storage policy to a file or a directory.
Command:
hdfs storagepolicies -setStoragePolicy -path <path> -policy <policy>
Arguments:
-path <path> | The path referring to either a directory or a file. |
-policy <policy> | The name of the storage policy. |
Unset a storage policy to a file or a directory. After the unset command the storage policy of the nearest ancestor will apply, and if there is no policy on any ancestor then the default storage policy will apply.
Command:
hdfs storagepolicies -unsetStoragePolicy -path <path>
Arguments:
-path <path> | The path referring to either a directory or a file. |